| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-09-27 | |||
| 20:55:48 | mriedem | cinder's db api code to get a volume filters on project_id https://github.com/openstack/cinder/blob/33b32d982094a9d85dabd737cb1d634b3164c187/cinder/db/sqlalchemy/api.py#L1893 | |
| 20:56:06 | mriedem | so if you're an admin with a different project trying to get the volume it's filtered out if that volume is in another project | |
| 20:56:13 | mriedem | i know you're familiar with that | |
| 20:59:12 | mriedem | idk how the hell volume show as admin works then... | |
| 21:21:15 | mriedem | ah because in devstack the admin and service projects have different roles, and the cinder user doesn't have the admin role | |
| 21:21:17 | mriedem | https://review.opendev.org/685488 | |
| 21:25:57 | efried | mriedem: I see a route for /versions/{id} but it's not documented. Does that mean I can rip it tf out? | |
| 21:26:34 | mriedem | probably not | |
| 21:26:56 | mriedem | lack of documentation or wrong documentation doesn't mean we can just remove things from the api since people in the wild might be using them, | |
| 21:27:01 | mriedem | and historically our api docs have sucked | |
| 21:29:02 | efried | FINE | |
| 21:29:43 | mriedem | sorry broheem! | |
| 21:29:49 | efried | mriedem: I think I figured out that paste thing I was asking about yesterday btw | |
| 21:30:02 | efried | https://review.opendev.org/685181 | |
| 21:30:37 | mriedem | i'm not really your guy for that, but alex_xu and/or johnthetubaguy might be | |
| 21:30:45 | mriedem | if either ever show up again | |
| 21:30:48 | efried | yah | |
| 21:31:08 | mriedem | just tell stephen that removing paste and using flask is on the table for ussuri | |
| 21:31:28 | efried | Funny, I was talking to lbragstad about this earlier and he showed me how they did that for keystone. | |
| 21:31:34 | mriedem | or whatever wsgi middleware thing is cool these days | |
| 21:31:46 | mriedem | what is neutron using now? pecan? | |
| 21:31:53 | efried | He said it took them two releases. I said I can't imagine how long that means it would take nova | |
| 21:32:03 | mriedem | oh neutron is still using paste | |
| 21:32:06 | efried | um, I thought I saw the same paste.ini snafu in neutron as we've got in nova... | |
| 21:32:10 | mriedem | i thought there was an effort there to move to pecan years ago | |
| 21:32:18 | efried | yeah, I was gonna say that may mean they're off of it and just haven't removed paste.ini yet. | |
| 21:32:27 | efried | that was part of keystone's process too. | |
| 21:32:38 | mriedem | idk when glance got off paste | |
| 21:32:54 | mriedem | removing paste form nova likely means breaking out of tree extension stuff and making a lot of people unhappy | |
| 21:32:59 | mriedem | i was just joking about doing it | |
| 21:33:50 | efried | I'm not laughing, but also realistic that it's never gonna happen. | |
| 21:34:02 | mriedem | ROI | |
| 21:34:09 | efried | roight. | |
| 21:34:18 | mriedem | dude, just submit it as a cycle goal | |
| 21:34:22 | mriedem | to de-paste openstack | |
| 21:34:25 | efried | ooo | |
| 21:34:28 | mriedem | people will just sign up to do it | |
| 21:35:30 | lbragstad | lol | |
| 21:36:38 | efried | see ^ | |
| 21:36:48 | efried | lbragstad is just itching to rip paste out of nova | |
| 21:36:57 | mriedem | no that's just a rash | |
| 21:37:00 | mriedem | he's rashy | |
| 21:37:04 | mriedem | diaper wise | |
| 21:37:17 | efried | razor burn | |
| 21:37:26 | mriedem | "rashy lance" we used to call him at ibm | |
| 21:37:49 | mriedem | with all of that helpfulness, i'm signing off, o/ | |
| 21:37:53 | efried | o/ | |
| 21:43:13 | openstackgerrit | Eric Fried proposed openstack/nova-specs master: Re-proposed Nova Cyborg interaction specification. https://review.opendev.org/682637 | |
| 22:08:44 | openstackgerrit | Merged openstack/nova-specs master: Re-proposed Nova Cyborg interaction specification. https://review.opendev.org/682637 | |
| 23:00:49 | openstackgerrit | Merged openstack/nova stable/rocky: lxc: make use of filter python3 compatible https://review.opendev.org/676498 | |
| #openstack-nova - 2019-09-28 | |||
| 00:00:27 | openstackgerrit | wangfaxin proposed openstack/nova master: fixd notes: neuton is replayced with neutron in line 756. https://review.opendev.org/685509 | |
| 02:02:22 | lbragstad | efried regarding the out-of-tree extension stuff... we broke that for keystone | |
| 02:03:30 | lbragstad | we formally pulled relevant middleware into the project and then added a few config options enable it | |
| 02:04:09 | lbragstad | otherwise - we punted on allowing people to put their own middleware in paste, kmalloc did all of that conversion for us | |
| 02:04:20 | lbragstad | TMMV | |
| 02:04:33 | lbragstad | s/TMMV/YMMV/ | |
| 07:24:03 | openstackgerrit | Merged openstack/nova master: Stop filtering out 'accepted' for in-progress migrations https://review.opendev.org/684409 | |
| 09:15:30 | openstackgerrit | OpenStack Proposal Bot proposed openstack/nova stable/train: Imported Translations from Zanata https://review.opendev.org/685534 | |
| #openstack-nova - 2019-09-29 | |||
| 01:54:19 | openstackgerrit | hulina proposed openstack/nova master: Fix spelling error https://review.opendev.org/685540 | |
| 09:00:29 | openstackgerrit | Luyao Zhong proposed openstack/nova master: make virtual pmem feature compatible with python3 https://review.opendev.org/685593 | |
| 09:04:59 | openstackgerrit | Luyao Zhong proposed openstack/nova master: make virtual pmem feature compatible with python3 https://review.opendev.org/685593 | |
| 23:35:47 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Update contributor guide for Ussuri https://review.opendev.org/685630 | |
| #openstack-nova - 2019-09-30 | |||
| 03:16:21 | openstackgerrit | Rui Zang proposed openstack/nova master: make virtual pmem feature compatible with python3 https://review.opendev.org/685593 | |
| 05:43:29 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Replace 'fake' with a real project ID https://review.opendev.org/685645 | |
| 05:58:54 | openstackgerrit | Rui Zang proposed openstack/nova master: make virtual pmem feature compatible with python3 https://review.opendev.org/685593 | |
| 07:23:51 | openstackgerrit | Lee Yarwood proposed openstack/nova master: DNM libvirt: extract _start_heterogeneous_computes https://review.opendev.org/682620 | |
| 07:23:55 | openstack | bug 1843639 in OpenStack Compute (nova) "libvirt: post_live_migration failures to disconnect volumes result in the rollback of live migrations" [Medium,In progress] https://launchpad.net/bugs/1843639 - Assigned to Lee Yarwood (lyarwood) | |
| 07:23:55 | openstackgerrit | Lee Yarwood proposed openstack/nova master: Add regression test for bug #1843639 https://review.opendev.org/682621 | |
| 07:23:59 | openstackgerrit | Lee Yarwood proposed openstack/nova master: libvirt: Ignore volume exceptions during post_live_migration https://review.opendev.org/682622 | |
| 07:25:16 | openstackgerrit | Lee Yarwood proposed openstack/nova master: compute: Remove stale BDMs on reserve_block_device_name failure https://review.opendev.org/682594 | |
| 07:29:37 | openstackgerrit | wangfaxin proposed openstack/nova master: fixd notes: neuton is replaced with neutron in line 756. https://review.opendev.org/685509 | |
| 09:00:03 | openstackgerrit | Seyeong Kim proposed openstack/nova master: block_device: Optionally recreate attachments when refreshing connection_info https://review.opendev.org/579004 | |
| 09:12:12 | openstackgerrit | Brin Zhang proposed openstack/nova-specs master: Fix invalid link index https://review.opendev.org/685664 | |
| 09:15:17 | bauzas | gibi: <3 you | |
| 09:15:24 | bauzas | gibi: found a shit number of helpers | |
| 09:15:35 | bauzas | in the integrated helpers for my functional tests | |
| 09:15:39 | bauzas | will make them easier to read | |
| 09:15:50 | gibi | bauzas: I'm glad that those can help you | |
| 09:20:33 | bauzas | gibi: the only issue is that I discovered those helpers after some time | |
| 09:21:07 | gibi | yeah it is hard how to make such (and other) helper codes discoverable | |
| 09:21:35 | bauzas | unless by reviewing others's patches :) | |
| 09:22:14 | gibi | but nobody can review every incoming patch in nova to know what happens | |
| 09:27:28 | bauzas | ...also if they have a lot of other things to do internally :) | |
| 09:29:54 | aspiers | Is it just me or is https://docs.openstack.org/glance/latest/admin/useful-image-properties.html wrong regarding the default of hw_scsi_model being virtio-blk? | |
| 09:31:36 | aspiers | AFAICS, if hw_scsi_model is not specified, there will be no SCSI controller | |
| 09:32:10 | aspiers | but also, "virtio-blk" is not a valid value in nova.objects.fields.SCSIModel | |
| 09:36:44 | aspiers | and in fact that doc is missing all the other valid values | |
| 10:16:03 | aspiers | ah, I think I figured it out | |
| 10:16:29 | openstackgerrit | Kashyap Chamarthy proposed openstack/nova master: libvirt: Use the `preferred` memory allocation mode for NUMA https://review.opendev.org/684375 | |
| 10:16:36 | aspiers | the default when hw_scsi_model is not specified is no SCSI controller, and something like <target dev='vda' bus='virtio'/> | |
| 10:17:03 | aspiers | which is colloquially referred to as virtio-blk, but it's just 'virtio' in the code | |
| 10:23:23 | sean-k-mooney | aspiers: yes the defualt for libvirt/kvm is virtio or virtio-blk or virtio block | |
| 10:23:58 | aspiers | sean-k-mooney: thanks for confirmation. just submitted https://review.opendev.org/#/c/685674/ | |
| 10:24:06 | sean-k-mooney | if you want to use virto-scsi you need to set the disk bus to scsi and then set the scsi model to virtio-scsi | |
| 10:25:16 | sean-k-mooney | the default storage contole depend on the hypervisor and virt type | |
| 10:25:32 | sean-k-mooney | it wont alwyas be virtio if you dont set hw_disk_bus | |
| 10:28:37 | sean-k-mooney | aspiers: this is the funciton in the libvirt driver that chooes the defult disk bus https://github.com/openstack/nova/blob/master/nova/virt/libvirt/blockinfo.py#L221-L296 | |
| 10:31:52 | sean-k-mooney | i should also point out that there is this horrible code that tries to incorrectly determin the optimal default disk and cd buss form the image os | |
| 10:31:54 | sean-k-mooney | https://github.com/openstack/nova/blob/master/nova/virt/osinfo.py#L114-L147 | |
| 10:33:16 | sean-k-mooney | but its actully broken by desgin in that it expect the OS version to be concatonated into the os_distro field instead of actully reading the os_version field so it does not actully work correctly | |